EA Challenge - part 2

List of required libraries.

rm(list=ls())

library(rio)
library(broom)
library(knitr)
library(plotly)
library(RColorBrewer)
library(kableExtra)
library(summarytools)
library(prettydoc)
library(jsonlite)
library(maps)
library(htmltools)
library(hrbrthemes)
library(DataExplorer)
library(tidyverse)
library(magrittr)
library(viridis)

DATA LOAD

mydata <-
    rio::import_list(dir("raw_data/",full.names = TRUE,pattern = "*.csv"),
                     rbind = FALSE)

# removing redundant data: "CompleteDataset"
mydata <- mydata[names(mydata) != "CompleteDataset"] 

# Force numeric all numeric columns in Player Attribute Data
mydata$PlayerAttributeData[]  <- lapply(mydata$PlayerAttributeData, function(x) as.numeric(as.character(x)))

# Remove index/rowname columns in all dataframes
mydata <- lapply(mydata, function(x) { x["V1"] <- NULL; x })

# Personal Data has an "Unnamed:0" Column. Get rid of it (First Col)
mydata$PlayerPersonalData <- mydata$PlayerPersonalData[2:ncol(mydata$PlayerPersonalData)]


# Extracting dataframes from list to global environment
# (This is optional. Working with list of objects is easier in R.
# Executed here just for easier workflow comprehension)
list2env(mydata, globalenv()) 

## <environment: R_GlobalEnv>

countries <- fromJSON("raw_data/countries.json") %>%
            data.table::rbindlist() %>%
            as_tibble() %>%
            mutate(languages = map(languages, ~ replace(.x, is.null(.x), "unknown"))) %>%  # avoid empty list items
            mutate(languages = unlist(languages)) %>%
            rename(lang_code = languages, country_name = name, native_name = native)



languages <- tibble(col = fromJSON("raw_data/languages.json")) %>%
             unnest_wider(col) %>%
             mutate(languages = names(fromJSON("raw_data/languages.json"))) %>%
             rename(lang_code = languages, lang_name = name, native_lang = native)



continents <- fromJSON("raw_data/continents.json") %>%
              unlist() %>%
              stack() %>%
              rename(continent = ind, continent_name = values)


final_languages_csv <-
              inner_join(countries, languages, by = "lang_code") %>%
              inner_join(continents, by = "continent") %>%
              select(-rtl)


players_languages <-
  inner_join(final_languages_csv,
             PlayerPersonalData,
             by = c("country_name" = "Nationality"))  %>%
  select(lang_name, Club, country_name, continent_name, Name) %>%
  group_by(Name) %>%
  summarise(
    languages = toString(unique(lang_name)),
    club =  first(Club),
    country = first(country_name),
    continent = first(continent_name)
  )

Players by Country

# Loading geographical data per player

players_by_country <- 
             PlayerPersonalData %>%
  group_by(Nationality) %>%
  tally() %>%
  magrittr::set_colnames(c("region","value"))%>% 
  ungroup() %>% 
   mutate(
    region = ifelse(region == "United States", "USA", region),
    region = ifelse(region == "DR Congo", "Democratic Republic of the Congo", region),
    region = ifelse(region == "China PR", "China", region),
    region = ifelse(region == "Congo", "Democratic Republic of the Congo", region),
    region = ifelse(region == "Bosnia Herzegovina", "Bosnia and Herzegovina", region),
    region = ifelse(region == "Scotland", "UK", region),
    region = ifelse(region == "England", "UK", region),
    region = ifelse(region == "Northern Ireland", "Ireland", region),
    region = ifelse(region == "Republic of Ireland", "Ireland", region),
    region = ifelse(region == "Central African Rep.", "Central African Republic", region),
    region = ifelse(region == "Wales", "UK", region)
    )  


# loading world maps data

world_map <- map_data("world")

# joining map data with players data

players_map <- left_join(players_by_country, world_map, by = "region") %>% 
  rename(players=value)

players_map$hover <- with(players_map, paste0(region,"<br>",players))

g <- ggplot(players_map, aes(long, lat, group = group, text=hover))+
  geom_polygon(aes(fill = players), color = "white")+
   scale_fill_viridis_c(option = "C") +
   labs(title='# Players by Continent / Country')+
  hrbrthemes::theme_ipsum() 


ggplotly(g,tooltip="text")

Top languages per Continent: Faceted version

continent_facets <- players_languages %>%
  group_by(continent,languages) %>% 
  tally() %>%
  arrange(desc(n)) %>% 
  top_n(n = 5) 


ggplot(continent_facets,
       aes(
         x = reorder(languages, n),
         y = n,
         fill = languages,
         palette = "dark2"
       )) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = n),
    vjust = 0.5,
    hjust = 1,
    color = "black",
    size = 4
  ) +
  xlab("") +
  coord_flip() +
  theme_minimal()+
  theme(
    legend.position = "none",
    plot.title = element_text(color = "black", size = 14, face = "bold"),
    plot.subtitle = element_text(size = 10)
  ) +
  facet_grid(~ continent, scales = "free", space = "free")

Languages per continent (non faceted version)

continent_facets %>% filter(continent=="Asia") %>% 
  ggplot( aes(
         x = reorder(languages, n),
         y = n,
         fill = languages,
         palette = "dark2"
       )) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = n),
    vjust = 0.5,
    hjust = 1,
    color = "white",
    size = 4
  ) +
  xlab("") +
  coord_flip() +
  theme_minimal()+
  theme(
    legend.position = "none",
    plot.title = element_text(color = "black", size = 14, face = "bold"),
    plot.subtitle = element_text(size = 10)
  ) +
  labs(title = "TOP 5 Asiatic Players Languages")

continent_facets %>% filter(continent=="Europe") %>% 
  ggplot( aes(
         x = reorder(languages, n),
         y = n,
         fill = languages,
         palette = "dark2"
       )) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = n),
    vjust = 0.5,
    hjust = 1,
    color = "white",
    size = 4
  ) +
  xlab("") +
  coord_flip() +
  theme_minimal()+
  theme(
    legend.position = "none",
    plot.title = element_text(color = "black", size = 14, face = "bold"),
    plot.subtitle = element_text(size = 10)
  ) +
  labs(title = "TOP 5 Europe Players Languages")

continent_facets %>% filter(continent=="Africa") %>% 
  ggplot( aes(
         x = reorder(languages, n),
         y = n,
         fill = languages,
         palette = "dark2"
       )) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = n),
    vjust = 0.5,
    hjust = 1,
    color = "white",
    size = 4
  ) +
  xlab("") +
  coord_flip() +
  theme_minimal()+
  theme(
    legend.position = "none",
    plot.title = element_text(color = "black", size = 14, face = "bold"),
    plot.subtitle = element_text(size = 10)
  ) +
  labs(title = "TOP 5 African players languages")

continent_facets %>% filter(continent=="South America") %>% 
  ggplot( aes(
         x = reorder(languages, n),
         y = n,
         fill = languages,
         palette = "dark2"
       )) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = n),
    vjust = 0.5,
    hjust = 1,
    color = "white",
    size = 4
  ) +
  xlab("") +
  coord_flip() +
  theme_minimal()+
  theme(
    legend.position = "none",
    plot.title = element_text(color = "black", size = 14, face = "bold"),
    plot.subtitle = element_text(size = 10)
  ) +
  labs(title = "TOP 5 South American players languages")

continent_facets %>% filter(continent=="North America") %>% 
  ggplot( aes(
         x = reorder(languages, n),
         y = n,
         fill = languages,
         palette = "dark2"
       )) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = n),
    vjust = 0.5,
    hjust = 1,
    color = "white",
    size = 4
  ) +
  xlab("") +
  coord_flip() +
  theme_minimal()+
  theme(
    legend.position = "none",
    plot.title = element_text(color = "black", size = 14, face = "bold"),
    plot.subtitle = element_text(size = 10)
  ) +
  labs(title = "TOP 5 North American players languages")

continent_facets %>% filter(continent=="Oceania") %>% 
  ggplot( aes(
         x = reorder(languages, n),
         y = n,
         fill = languages,
         palette = "dark2"
       )) +
  geom_bar(stat = "identity") +
  geom_text(
    aes(label = n),
    vjust = 0.5,
    hjust = 1,
    color = "white",
    size = 4
  ) +
  xlab("") +
  coord_flip() +
  theme_minimal()+
  theme(
    legend.position = "none",
    plot.title = element_text(color = "black", size = 14, face = "bold"),
    plot.subtitle = element_text(size = 10)
  ) +
  labs(title = "TOP 5 Oceania players languages")